Nej Mo A 1300955

n engl j med 371;3 nejm.
org july 17, 2014

203
The new england
journal of medicine
established in 1812 july 17, 2014 vol. 371 no. 3
Effects of Extended-Release Niacin with Laropiprant
in High-Risk Patients
The HPS2-THRIVE Collaborative Group*
ABSTRACT
The members of the writing committee
(Martin J. Landray, Ph.D., F.R.C.P., Richard
Haynes, D.M., M.R.C.P., Jemma C. Hope-
well, Ph.D., Sarah Parish, D.Phil., Theingi
Aung, M.B., B.S., M.R.C.P., Joseph Tomson,
M.R.C.P., Karl Wallendszus, M.Sc., and
Martin Craig, Ph.D., Clinical Trial Service
Unit, University of Oxford, Oxford, United
Kingdom; Lixin Jiang, M.D., Ph.D., Cardio-
vascular Institute and Fuwai Hospital, Chi-
nese Academy of Medical Sciences, Beijing;
and Rory Collins, F.Med.Sci., F.R.C.P.,
and Jane Armitage, F.R.C.P, F.F.P.H., Clini-
cal Trial Service Unit, University of Oxford,
Oxford, United Kingdom) assume respon-
sibility for the content and integrity of this
article. Address reprint requests to Dr.
Armitage at the Clinical Trial Service Unit
and Epidemiological Studies Unit, Nuffield
Department of Population Health, Richard
Doll Bldg., Old Road Campus, Roosevelt Dr.,
Oxford OX3 7LF, United Kingdom, or at
jane.armitage@ctsu.ox.ac.uk.
* A complete list of collaborators in the
Heart Protection Study 2Treatment of
HDL to Reduce the Incidence of Vascu-
lar Events (HPS2-THRIVE) is provided
in the Supplementary Appendix, avail-
able at NEJM.org.
N Engl J Med 2014;371:203-12.
DOI: 10.1056/NEJMoa1300955
Copyright 2014 Massachusetts Medical Society.
BACKGROUND
Patients with evidence of vascular disease are at increased risk for subsequent vascular
events despite effective use of statins to lower the low-density lipoprotein (LDL)
cholesterol level. Niacin lowers the LDL cholesterol level and raises the high-density
lipoprotein (HDL) cholesterol level, but its clinical efficacy and safety are uncertain.
METHODS
After a prerandomization run-in phase to standardize the background statin-based LDL
cholesterollowering therapy and to establish participants ability to take extended-
release niacin without clinically significant adverse effects, we randomly assigned
25,673 adults with vascular disease to receive 2 g of extended-release niacin and 40 mg
of laropiprant or a matching placebo daily. The primary outcome was the first major
vascular event (nonfatal myocardial infarction, death from coronary causes, stroke,
or arterial revascularization).
RESULTS
During a median follow-up period of 3.9 years, participants who were assigned to
extended-release niacinlaropiprant had an LDL cholesterol level that was an average
of 10 mg per deciliter (0.25 mmol per liter as measured in the central laboratory)
lower and an HDL cholesterol level that was an average of 6 mg per deciliter (0.16
mmol per liter) higher than the levels in those assigned to placebo. Assignment to
niacinlaropiprant, as compared with assignment to placebo, had no significant effect
on the incidence of major vascular events (13.2% and 13.7% of participants with an
event, respectively; rate ratio, 0.96; 95% confidence interval [CI], 0.90 to 1.03; P = 0.29).
Niacinlaropiprant was associated with an increased incidence of disturbances in dia-
betes control that were considered to be serious (absolute excess as compared with
placebo, 3.7 percentage points; P<0.001) and with an increased incidence of diabetes
diagnoses (absolute excess, 1.3 percentage points; P<0.001), as well as increases in
serious adverse events associated with the gastro intestinal system (absolute excess, 1.0
percentage point; P<0.001), musculoskeletal system (absolute excess, 0.7 percentage
points; P<0.001), skin (absolute excess, 0.3 percentage points; P = 0.003), and unexpect-
edly, infection (absolute excess, 1.4 percentage points; P<0.001) and bleeding (absolute
excess, 0.7 percentage points; P<0.001).
CONCLUSIONS
Among participants with atherosclerotic vascular disease, the addition of extended-
release niacinlaropiprant to statin-based LDL cholesterollowering therapy did not
significantly reduce the risk of major vascular events but did increase the risk of seri-
ous adverse events. (Funded by Merck and others; HPS2-THRIVE ClinicalTrials.gov
number, NCT00461630.)
The New England Journal of Medicine
Downloaded from nejm.org on July 18, 2014. For personal use only. No other uses without permission.
Copyright 2014 Massachusetts Medical Society. All rights reserved.
The new engl and journal o f medicine
n engl j med 371;3 nejm.org july 17, 2014
204
P
atients with cardiovascular disease
remain at substantial risk for major vascular
events despite current approaches to treat-
ment of risk factors.
1
Observational data indicate
that the low-density lipoprotein (LDL) cholesterol
level is strongly positively associated with the risk
of coronary heart disease and that the high-density
lipoprotein (HDL) cholesterol level is strongly in-
versely associated.
2
High-dose niacin decreases
the LDL cholesterol level and increases the HDL
cholesterol level, as well as lowering triglyceride
and lipoprotein(a) levels and blood pressure.
3,4

Current guidelines recommend that niacin therapy
be considered for reducing cardiovascular risk,
5,6

and its use in the United States has been increasing
steadily,
7
despite the lack of evidence from random-
ized trials of a clinical benefit when niacin is added
to current treatment.
The Atherothrombosis Intervention in Meta-
bolic Syndrome with Low HDL/High Triglycerides:
Impact on Global Health Outcomes (AIM-HIGH)
trial, which involved 3414 high-risk patients who
were receiving statin therapy, was stopped pre-
maturely after 3 years because of an apparent
lack of benefit with extended-release niacin.
8

However, given the small differences in blood
lipid levels that were observed between random-
ized groups, the AIM-HIGH trial may have been
too small to detect plausible reductions in vascu-
lar events.
The Heart Protection Study 2Treatment of
HDL to Reduce the Incidence of Vascular Events
(HPS2-THRIVE) was designed to assess the ef-
fects of adding extended-release niacin in com-
bination with laropiprant to effective statin-based
LDL cholesterollowering treatment in 25,673 high-
risk patients with prior vascular disease. Laro-
piprant is an antagonist of the prostaglandin D2
receptor DP
1
that has been shown to improve
adherence to niacin therapy by reducing flushing
in up to two thirds of patients.
9,10
METHODS
STUDY ORGANIZATION
HPS2-THRIVE was a randomized, double-blind,
multicenter trial that enrolled patients at 245 sites
in the United Kingdom (89 sites), Scandinavia
(84), and China (72). The trial was designed, con-
ducted, and analyzed by the Clinical Trial Service
Unit at Oxford University, which was the inde-
pendent regulatory sponsor of the trial. Merck
(manufacturer of the study drugs) funded the
trial, had nonvoting membership on the steering
committee, and provided trial coordination with-
in Scandinavia through its subsidiaries (under the
direction of Oxford University). Although Merck
had the opportunity to comment on preliminary
drafts of the manuscript, it otherwise had no role
in the design or conduct of the trial, the analysis
of the data, the approval of the manuscript, or
the decision to submit it for publication. The
study protocol (available with the full text of this
article at NEJM.org) was approved by the relevant
institutional review board for each participating
center. The last member of the writing commit-
tee vouches for the data and analyses and for the
fidelity of this report to the study protocol.
As prespecified, the trial design, characteristics
of the study participants, and effects of niacin
laropiprant on certain safety outcomes were re-
ported before the scheduled end of the study.
11

The methods are summarized below, with fur-
ther details provided in Supplementary Appen-
dix 1, available at NEJM.org.
STUDY PARTICIPANTS
Men and women 50 to 80 years of age were eli-
gible if they had a history of myocardial infarc-
tion, cerebrovascular disease, peripheral arterial
disease, or diabetes mellitus with evidence of
symptomatic coronary disease. There were no
entry criteria regarding lipid levels. Patients were
excluded if they had clinically significant hepat-
ic, renal, muscle-related, or other disease, were
receiving concurrent treatment with potentially
interacting drugs, or were receiving LDL choles-
terollowering treatment that was more effective
than simvastatin at a dose of 40 mg plus ezetimibe
at a dose of 10 mg daily.
11
Details of the trial
inclusion and exclusion criteria are provided in
Supplementary Appendix 1.
STUDY PROCEDURES
Potentially eligible patients were identified from
hospital or clinic records, or by means of adver-
tisement, and were invited to attend a study clinic.
Patients who appeared to be eligible for inclusion
in the study were asked to provide written in-
formed consent and to discontinue any statin
therapy. In a prerandomization run-in phase,
each participant received simvastatin at a dose of
40 mg daily; if this dose was not as effective as
their prior treatment or if their total cholesterol
Niacin with Laropiprant in High-Risk Patients
205
level was 135 mg per deciliter (3.50 mmol per liter)
or higher after 4 weeks, ezetimibe at a dose of
10 mg daily was added. After LDL cholesterol
lowering therapy had been standardized, partici-
pants received a combination tablet containing
1 g of extended-release niacin and 20 mg of laro-
piprant daily for 4 weeks, followed by two tablets
daily providing a total of 2 g of niacin and 40 mg
of laropiprant for 3 to 6 weeks. Participants who
did not report clinically significant adverse effects
with this treatment and who remained eligible
11

were randomly assigned to receive two niacin
laropiprant combination tablets (a total of 2 g
of niacin and 40 mg of laropiprant) daily or
matching placebo.
After randomization, follow-up assessments
of participants were to take place at 3 months
and 6 months and then every 6 months for a
median duration of 4 years. All serious adverse
events, including potential study outcomes, were
recorded. In addition, nonserious adverse events
that were considered by the participants to be
related to the study drug or that resulted in dis-
continuation of the study drug were recorded, as
were symptoms of muscle pain or weakness and
hepatitis (e.g., nausea, vomiting, or jaundice). Ad-
herence, which was defined as self-reported con-
sumption of at least 80% of the study drug, was
recorded, and blood tests were performed for liver
and muscle safety monitoring.
STUDY OUTCOMES
The primary outcome was the first major vascu-
lar event, defined as a major coronary event
(nonfatal myocardial infarction or death from
coronary causes), stroke of any type, or coronary
or noncoronary revascularization. Secondary out-
comes included the components of the primary
outcome, different types of stroke, and mortality
(overall and in specific categories). Additional
prespecified secondary outcomes included the
primary outcome after the exclusion of hemor-
rhagic stroke, the primary outcome after the ex-
clusion of both hemorrhagic stroke and any arte-
rial revascularization procedure, and the primary
outcome separately in the first year and in later
years. Detailed definitions of the outcomes for
the prespecified analyses are provided in Supple-
mentary Appendix 1.
All the reports of possible major vascular
events or safety outcomes were centrally adjudi-
cated according to prespecified criteria by clini-
cians who were unaware of the study-treatment
assignments. Analyses were based on confirmed
plus unrefuted reports of events; 95 to 99% of
the myocardial infarctions, strokes, and revascu-
larizations that were included were confirmed.
STATISTICAL ANALYSIS
On the basis of data from trials involving similar
populations
12-14
and expected secular trends, we
calculated that 20,000 participants would need
to undergo randomization for the study to have
more than 95% power to detect a proportional
reduction in risk of 15% in the primary outcome
with niacinlaropiprant, at a significance level of
less than 0.05 (see Supplementary Appendix 1).
However, on the basis of the observed changes in
the lipid levels during the prerandomization run-in
phase and adherence to the regimen during the
scheduled randomized study-treatment period,
it was estimated that differences in the lipid
levels during the randomized phase would be
smaller than originally anticipated. Consequent-
ly, the steering committee increased the sample
to 25,000 participants; we estimated that with
more than 3400 major vascular events, the study
would have more than 80% power to detect a
proportional reduction in risk of 10%, at a sig-
nificance level of less than 0.05.
Data-analysis plans were prespecified in the
original protocol and published on the study web-
site (www.ctsu.ox.ac.uk/thrive) before unblinded
analyses were available to the steering commit-
tee. Prespecified comparisons involved log-rank
analyses of the first occurrence of particular
events during the scheduled treatment period
after randomization among all the participants
assigned to niacinlaropiprant versus all those
assigned to placebo (i.e., intention-to-treat analy-
ses).
15,16
The log-rank analysis yielded the average
rate ratio for an event or death, with the pro-
portional reduction in this ratio expressed as a
percentage, and a two-sided significance test.
RESULTS
STUDY PARTICIPANTS
Of 51,698 people who were screened, 42,424 en-
tered the prerandomization run-in phase (Fig. S1
in Supplementary Appendix 1). In summary,
11.2% of the patients who started the LDL cho-
lesterolstandardization phase withdrew during
this phase, and 33.1% of those who started the
206
active niacinlaropiprant phase withdrew during
that phase (chiefly because of skin-related, gastro-
intestinal, diabetes-related, and musculoskeletal
adverse effects).
11
From April 2007 through July
2010, a total of 25,673 participants underwent
randomization (Table 1, and Fig. S1 in Supple-
mentary Appendix 1). The mean age of the par-
ticipants was 64.9 years, and 82.7% of the partici-
pants were men.
Coronary disease was reported by 78.4% of the
participants, cerebrovascular disease by 31.8%,
peripheral arterial disease by 12.5%, and diabetes
mellitus by 32.3%, and 36.3% of the participants
met the criteria for the metabolic syndrome.
11,17

Baseline cholesterol levels were well controlled with
the LDL cholesterollowering regimen established
during the run-in phase, with an average LDL cho-
lesterol level of 63 mg per deciliter (1.64 mmol
per liter, as measured in the central laboratory)
and an average HDL cholesterol level of 44 mg per
deciliter (1.14 mmol per liter). The median dura-
tion of follow-up was 3.9 years (mean, 3.6 years),
yielding 46,239 person-years in the niacinlaro-
piprant group and 46,359 person-years in the
placebo group.
ADHERENCE AND LIPID LEVELS
Reported adherence to niacinlaropiprant fell to
89.1% during the first year and to 69.9% by the
scheduled end of the study period, yielding a
study average of 77.7% (Table S1 in Supplemen-
tary Appendix 1). Overall, significantly more par-
ticipants in the niacinlaropiprant group than in
the placebo group discontinued the study drug
(25.4% vs. 16.6%, P<0.001), with discontinuations
attributable primarily to recognized side effects
of niacin (e.g., skin-related, gastrointestinal, mus-
culoskeletal, and diabetes-related adverse events).
11

Participants who discontinued the randomized
study drug were also more likely to discontinue
their background-study LDL cholesterollowering
treatment and were encouraged to take a nonstudy
statin. There was negligible use of nonstudy niacin
(three participants in the niacinlaropiprant group
and nine in the placebo group).
During the study, assignment to treatment with
niacinlaropiprant was associated with an average
reduction in the LDL cholesterol level of 10 mg
per deciliter (0.25 mmol per liter, as measured in
the central laboratory), an average increase in
the HDL cholesterol level of 6 mg per deciliter
(0.16 mmol per liter), and an average reduction
in the triglyceride level of 33 mg per deciliter
(0.37 mmol per liter), as compared with assign-
ment to placebo (Tables S2 and S3 in Supplemen-
tary Appendix 1). Niacinlaropiprant was also
associated with a number of other effects, in-
cluding reductions in weight, blood pressure,
and lipoprotein(a) level and an increase in the
glycated hemoglobin level (Table S4 in Supple-
mentary Appendix 1).
EFFECTS ON MAJOR VASCULAR EVENTS
Assignment to niacinlaropiprant, as compared
with assignment to placebo, was not associated
with a significant reduction in the incidence of
major vascular events (1696 participants with
events [13.2%] and 1758 participants with events
[13.7%], respectively; rate ratio, 0.96; 95% confi-
dence interval [CI], 0.90 to 1.03; P = 0.29) (Fig. 1
and 2). There was no apparent effect in the first
year after randomization (rate ratio, 1.01; 95%
CI, 0.90 to 1.14) or in subsequent years (rate ra-
tio, 0.94; 95% CI, 0.87 to 1.02; P = 0.17). Similarly,
there were no significant effects of assignment to
niacinlaropiprant, as compared with assignment
to placebo, on the secondary outcomes of the
incidence of major vascular events excluding
hemorrhagic stroke (12.4% and 13.1%, respec-
tively; rate ratio, 0.95; 95% CI, 0.88 to 1.01;
P = 0.12) or excluding both hemorrhagic stroke
and revascularization procedures (7.9% and
8.4%, respectively; rate ratio, 0.95; 95% CI, 0.87
to 1.03; P = 0.20).
With respect to the separate components of
major vascular events, there was no significant
effect of niacinlaropiprant, as compared with
placebo, on the incidence of major coronary
events (rate ratio, 0.96; 95% CI, 0.87 to 1.07;
P = 0.51) or any stroke (rate ratio, 1.00; 95% CI,
0.88 to 1.13; P = 0.56), but there was a nominally
significant 10% proportional reduction in arteri-
al revascularization procedures (rate ratio, 0.90;
95% CI, 0.82 to 0.99; P = 0.03). With respect to
subtypes of stroke, there was no significant effect
of niacinlaropiprant as compared with placebo on
the incidence of presumed ischemic stroke (3.0%
and 3.2%, respectively; rate ratio, 0.94; 95% CI,
0.82 to 1.08) or hemorrhagic stroke (0.9% vs. 0.7%,
respectively; rate ratio, 1.28; 95% CI, 0.97 to 1.69).
SUBGROUP ANALYSES
Prespecified secondary analyses examined the ef-
fect of niacinlaropiprant on the incidence of ma-
jor vascular events in subgroups defined according
to history of various types of vascular disease or
207
diabetes. Between the subgroups, there were no
significant differences in the absolute changes in
lipid levels or the proportional reductions in risk
(Table S3 and Fig. S2 in Supplementary Appendix 1).
Tertiary assessments included the effects on
major vascular events in more than 30 additional
prespecified subgroup categories. Even without
adjustment for multiple comparisons, there were
no significant differences among most of these
subgroups in the proportional reductions in ma-
Table 1. Characteristics of the Participants at Baseline.*
Characteristic
NiacinLaropiprant
(N = 12,838)
Placebo
(N = 12,835)
All Participants
(N = 25,673)
Age
Distribution no. (%)
<65 yr 6,470 (50.4) 6,462 (50.3) 12,932 (50.4)
65 to <70 yr 2,814 (21.9) 2,810 (21.9) 5,624 (21.9)
70 yr 3,554 (27.7) 3,563 (27.8) 7,117 (27.7)
Mean yr 64.97.5 64.97.5 64.97.5
Sex no. (%)
Male 10,614 (82.7) 10,615 (82.7) 21,229 (82.7)
Female 2,224 (17.3) 2,220 (17.3) 4,444 (17.3)
Region no. (%)
Europe 7,374 (57.4) 7,367 (57.4) 14,741 (57.4)
China 5,464 (42.6) 5,468 (42.6) 10,932 (42.6)
Prior disease no. (%)
Myocardial infarction 8,685 (67.7) 8,660 (67.5) 17,345 (67.6)
Other coronary heart disease 8,424 (65.6) 8,346 (65.0) 16,770 (65.3)
Cerebrovascular disease 4,058 (31.6) 4,112 (32.0) 8,170 (31.8)
Peripheral arterial disease 1,622 (12.6) 1,592 (12.4) 3,214 (12.5)
Diabetes 4,134 (32.2) 4,165 (32.5) 8,299 (32.3)
LDL cholesterol
<58 mg/dl 4,933 (38.4) 4,927 (38.4) 9,860 (38.4)
58 to <77 mg/dl 5,505 (42.9) 5,549 (43.2) 11,054 (43.1)
77 mg/dl 2,400 (18.7) 2,359 (18.4) 4,759 (18.5)
Mean mg/dl 6417 6317 6317
HDL cholesterol
<35 mg/dl 2,459 (19.2) 2,441 (19.0) 4,900 (19.1)
35 to <43 mg/dl 4,098 (31.9) 4,037 (31.5) 8,135 (31.7)
43 mg/dl 6,281 (48.9) 6,357 (49.5) 12,638 (49.2)
Mean mg/dl 43.911.2 44.011.2 43.911.2
Total cholesterol
<116 mg/dl 3,547 (27.6) 3,582 (27.9) 7,129 (27.8)
116 to <135 mg/dl 4,800 (37.4) 4,691 (36.5) 9,491 (37.0)
135 mg/dl 4,491 (35.0) 4,562 (35.5) 9,053 (35.3)
Mean mg/dl 12822 12822 12822
* Plusminus values are means SD. There were no significant differences in the baseline characteristics between the
study groups. To convert the values for cholesterol to millimoles per liter, multiply by 0.02586. Additional details are
provided in Figure S2 in Supplementary Appendix 1.
208
jor vascular events (Fig. S2 in Supplementary
Appendix 1). An unadjusted P value of less than
0.05 for heterogeneity or trend was observed in
the following categories: smoking status, alco-
hol intake, beta-blocker use, and LDL cholesterol
and apolipoprotein B levels. The nominally sig-
nificant trend (P = 0.02) toward a greater reduc-
tion in risk in the subgroup with a higher base-
line LDL cholesterol level may be related, at least
in part, to the greater reduction in the LDL cho-
lesterol level in that subgroup. Similarly, apparent
differences between other subgroups in the ob-
served effects on major vascular events may be
related to differences between them in observed
changes in the lipid levels, or they may be due to
chance (Table S3 in Supplementary Appendix 1).
Niacin has been recommended particularly
for patients who have the combination of low
HDL cholesterol and high triglyceride levels.
5,6

Consequently, an exploratory analysis was con-
ducted that included participants with both an
HDL cholesterol level of less than 40 mg per
deciliter (1.03 mmol per liter) in men or 51 mg
per deciliter (1.32 mmol per liter) in women and a
triglyceride level of more than 151 mg per deci-
liter (1.70 mmol per liter) in either sex. In this
analysis, there was no significant difference be-
tween the niacinlaropiprant group and the
placebo group in the incidence of major vascu-
lar events (333 of 2203 participants [15.1%] and
334 of 2159 participants [15.5%], respectively,
with an event).
EFFECTS ON MORTALITY AND RATES OF CANCER
Niacinlaropiprant, as compared with placebo,
was associated with a nonsignificant 9% propor-
tional increase in the incidence of death from any
cause (798 participants [6.2%] and 732 partici-
pants [5.7%], respectively; rate ratio, 1.09; 95% CI,
0.99 to 1.21; P = 0.08), with similar nonsignificant
increases in both vascular and nonvascular mor-
tality (Fig. S3 in Supplementary Appendix 1). No
significant increases were seen in any prespeci-
fied subgroups with regard to specific causes of
death, including cancer. There were no signifi-
cant increases in the incidence of cancer overall
(4.8% with niacinlaropiprant and 4.7% with
placebo, P = 0.67) or at any prespecified site (Fig.
S4 in Supplementary Appendix 1).
effects on other ADVERSE EVENTS
Assignment to niacinlaropiprant, as compared
with assignment to placebo, was associated with a
highly significant excess of participants with fatal
or nonfatal serious adverse events (7137 [55.6%]
vs. 6762 [52.7%], P<0.001), with many partici-
pants having more than one serious adverse event
(Table S5 in Supplementary Appendix 1). The
largest excesses in serious adverse events (non-
fatal and fatal combined) were related to effects
on glucose metabolism (Table 2, and Table S6 in
Supplementary Appendix 1). In an analysis of the
8299 participants who had diabetes at the time of
randomization, assignment to niacinlaropiprant,
as compared with assignment to placebo, was as-
sociated with a 55% proportional increase in dis-
turbances in diabetes control that were considered
to be serious, most of which led to hospitalization
(11.1% vs. 7.5%, P<0.001). In an analysis of the
17,374 participants who did not have diabetes at
the time of randomization, assignment to niacin
laropiprant, as compared with assignment to pla-
cebo, was associated with a 32% proportional in-
crease in the diagnosis of diabetes (5.7% vs. 4.3%,
P<0.001).
P
a
r
t
i
c
i
p
a
n
t
s

w
i
t
h

E
v
e
n
t

(
%
)
100
80
90
70
60
40
30
10
50
20
0
0 1 2 3 4
Years of Follow-up
P=0.29 by log-rank test
No. at Risk
Niacinlaropiprant
Placebo
Benefit per 1000
participants
assigned to
niacinlaropiprant
12,838
12,835
12,232
12,247
03
11,517
11,523
33
7672
7643
55
4978
5036
57
Placebo
Niacin
laropiprant
20
10
5
15
0
0 1 2 3 4
15.0
14.5
Figure 1. First Major Vascular Event during Follow-up.
Shown are KaplanMeier plots of the first major vascular event during the
4 years of follow-up. The inset shows the same data on an expanded y axis.
The numbers of participants at risk for a first postrandomization major vas-
cular event at the start of each year of follow-up are also shown, along with
the benefit, which is shown as the absolute differences (with standard errors)
in incidence rates between participants assigned to niacinlaropiprant and
those assigned to placebo.
209
There were also highly significant excesses of
other recognized adverse effects of niacin, includ-
ing gastrointestinal, musculoskeletal, and skin-
related serious adverse events. The excess of
gastrointestinal serious adverse events in the
niacinlaropiprant group as compared with the
placebo group (4.8% vs. 3.8%, P<0.001) included
bleeding and peptic ulceration, as well as other
problems in the upper and lower gastrointestinal
tracts (mostly dyspepsia and diarrhea, respec-
tively). The excess of musculoskeletal serious
adverse events with niacinlaropiprant (3.7%
vs. 3.0%, P<0.001) reflected primarily a risk of
myop athy with niacinlaropiprant that was four
times as high as that with placebo,
11
plus a
smaller excess of gout. In contrast to the large
excess in skin-related nonserious adverse events
that led to some participants discontinuing nia-
cinlaropiprant,
11
there was only a small excess
of skin-related serious adverse events (0.7% vs.
0.4%, P = 0.003) mostly rashes and skin ulcer-
ations.
In addition to these known side effects of nia-
cin, niacinlaropiprant was associated with highly
significant excesses of infection and bleeding
that were considered to be serious (Table 2). The
excess of infections with niacinlaropiprant ver-
sus placebo (8.0% vs. 6.6%, P<0.001) was dis-
tributed across a range of sites and types of in-
fection (Table S6 in Supplementary Appendix 1).
Similarly, the excess of bleeding events with niacin
laropiprant versus placebo (2.5% vs. 1.9%, P<0.001)
was distributed across a number of sites (and in-
cluded gastrointestinal bleeding and intracranial
hemorrhage).
Searchable tabulations of all the serious ad-
verse events (fatal and nonfatal combined), as well
as all the adverse reactions that were not consid-
ered to be serious, are provided in Supplemen-
tary Appendix 2, available at NEJM.org. They are
grouped on the basis of the Medical Dictionary for
Regulatory Activities, version 14.0, classification
system, according to system organ class, higher-
level general term, and higher-level term.
0.8 1.0 1.5
Placebo
Better
NiacinLaropiprant
Better
Major coronary event
Nonfatal myocardial infarction
Death from coronary cause
Any major coronary event
Stroke
Nonhemorrhagic stroke
Hemorrhagic stroke
Any stroke
Revascularization procedure
Coronary revascularization
Noncoronary revascularization
Any revascularization procedure
Any major vascular event
NiacinLaropiprant
(N=12,838) Rate Ratio (95% CI)
Placebo
(N=12,835) Type of Major Vascular Event
0.96 (0.901.03)
0.90 (0.820.99)
1.00 (0.881.13)
0.96 (0.871.07)
P Value
402 (3.1)
302 (2.4)
668 (5.2)
389 (3.0)
114 (0.9)
498 (3.9)
591 (4.6)
236 (1.8)
807 (6.3)
1696 (13.2)
431 (3.4)
291 (2.3)
694 (5.4)
415 (3.2)
89 (0.7)
499 (3.9)
664 (5.2)
258 (2.0)
897 (7.0)
1758 (13.7)
0.51
0.56
0.03
0.29
no. of participants with event (%)
Figure 2. First Major Vascular Event, According to Type of Event.
Counts (and percentages) are of the participants with a first event of the listed type during follow-up. Since a single participant may have
had multiple events, there is some nonadditivity between different types of event. Rate ratios comparing the outcome among partici-
pants assigned to niacinlaropiprant with the outcome among those assigned to placebo are plotted. For subcategories, rate ratios are
plotted as squares, with the size of each square proportional to the amount of statistical information, and horizontal lines represent
95% confidence intervals. For subtotals and totals, rate ratios and corresponding 95% confidence intervals are represented by diamonds,
with the rate ratios, 95% confidence intervals, and statistical significance tests given alongside. Squares or diamonds to the left of the
solid vertical line indicate benefit with niacinlaropiprant, but the benefit is significant (P<0.05) only if the horizontal line or diamond
does not overlap the solid vertical line. The overall rate ratio is indicated by the dashed vertical line.
210
DISCUSSION
HPS2-THRIVE showed that adding niacinlaro-
piprant to effective statin-based LDL cholesterol
lowering therapy in patients known to have vascu-
lar disease did not significantly reduce the risk of
major vascular events, either overall or in any
particular subgroup of patients. However, the study
also identified significant hazards, some of which
had not been reported previously with niacin.
On the basis of meta-analyses of statin trials,
the reduction of 10 mg per deciliter in the LDL
cholesterol level in HPS2-THRIVE would have been
expected to produce a 5 to 6% proportional re-
duction in the risk of major vascular events.
1

There is no similar evidence from randomized
trials regarding the effects of raising the HDL
cholesterol level, but if the inverse association
with vascular disease risk in observational stud-
ies is causal (which has been questioned)
18
and
half of it is reversible within a few years (as with
LDL cholesterollowering therapy
1
) then the in-
crease of 6 mg per deciliter in the HDL choles-
terol level observed in HPS2-THRIVE might have
been associated with a reduction in the risk of
major vascular events of 4 to 5%.
2
Consequently,
the combined changes in the lipid levels would
have been expected to reduce the risk of major
vascular events by approximately 10%, which is
slightly larger than the observed result (although
still statistically compatible with it).
The findings regarding major vascular events in
HPS2-THRIVE are consistent with those of previous
randomized trials of high-dose niacin alone. In the
Coronary Drug Project (CDP), which was conduct-
ed before effective LDL cholesterollowering agents
were available, niacin reduced the total cholesterol
level by 26 mg per deciliter (0.67 mmol per liter)
from a high baseline level of 253 mg per deciliter
(6.54 mmol per liter).
19
Lipid fractions were not
measured in the CDP, but it can be estimated that
the LDL cholesterol level was reduced by at least
30 mg per deciliter (0.78 mmol per liter) and the
HDL cholesterol level was increased by approxi-
mately 5 mg per deciliter (0.13 mmol per liter).
On the basis of the statin trials and observa-
tional epidemiologic studies,
1,2
such changes in
lipid levels might produce a reduction in risk of
15 to 20%, which is compatible with the 19%
reduction in myocardial infarction or coronary
death observed in the CDP. By contrast, in the
AIM-HIGH trial,
8
adding niacin to effective LDL
Table 2. Effects of NiacinLaropiprant on Selected Serious Adverse Events and Diabetes.*
Event Type
Niacin
Laropiprant
(N = 12,838)
Placebo
(N = 12,835)
Rate Ratio
(95% CI)
Absolute Excess
with Niacin
Laropiprant P Value
percentage points
Serious adverse event no. (%)
Gastrointestinal event 620 (4.8) 491 (3.8) 1.28 (1.131.44) 1.00.3 <0.001
Musculoskeletal event 481 (3.7) 385 (3.0) 1.26 (1.101.44) 0.70.2 <0.001
Skin-related event 86 (0.7) 51 (0.4) 1.67 (1.202.34) 0.30.1 0.003
Infection event 1031 (8.0) 853 (6.6) 1.22 (1.121.34) 1.40.3 <0.001
Bleeding event 326 (2.5) 238 (1.9) 1.38 (1.171.62) 0.70.2 <0.001
Diabetes mellitus no./total no. (%)
New-onset diabetes in participants
without diabetes at baseline
494/8704 (5.7) 376/8670 (4.3) 1.32 (1.161.51) 1.30.3 <0.001
Disturbed diabetes control in
participants with diabetes
at baseline
460/4134 (11.1) 311/4165 (7.5) 1.55 (1.341.78) 3.70.6 <0.001
* Plusminus values are means SE. Results are shown for the serious adverse events (fatal and nonfatal combined) for which there was a
significant difference between the randomized study groups. All the categories are mutually exclusive, except for bleeding, which includes
serious adverse events from various categories. Only participants with diabetes at randomization were at risk for disturbed diabetes control,
and only those without diabetes at randomization were at risk for new-onset diabetes (defined here as either self-reported new-onset diabe-
tes or new use of medication for glycemic control). In an analysis that included patients with an elevated glycated hemoglobin level (defined
as >48 mmol of glycated hexapeptide per mole of total hexapeptide, according to the method recommended by the International Federation
of Clinical Chemistry and Laboratory Medicine) in the definition of new-onset diabetes, we found 792 cases (9.1%) with niacinlaropiprant
versus 632 (7.3%) with placebo (rate ratio, 1.27 [95% CI, 1.14 to 1.41]; absolute excess, 1.80.4 percentage points; P<0.001). Additional de-
tails are provided in Table S6 in Supplementary Appendix 1.
211
cholesterollowering therapy reduced the LDL
cholesterol level by only 5 mg per deciliter and
increased the HDL cholesterol level by 5 mg per
deciliter. Such changes in lipid levels would be
expected to reduce the relative risk of major
vascular events by less than 10%, which is also
compatible with the observed result in that trial.
In HPS2-THRIVE, niacinlaropiprant was as-
sociated with highly significant increases in the
rates of various serious adverse events, including
some already known to be caused by niacin (i.e.,
diabetes-related, gastrointestinal, musculoskeletal,
and skin-related disorders).
8,19,20
For example, new
diagnoses of diabetes were increased by one third,
corresponding to 13 new cases per 1000 patients
treated for approximately 4 years (or 18 new cases
per 1000 patients if a glycated hemoglobin level
of more than 48 mmol per mole is included in
the definition). In addition, among the partici-
pants with diabetes at baseline, serious compli-
cations associated with glucose control (most of
which resulted in hospitalization) occurred in
37 patients per 1000, a finding that counters
previous reassurances about the safety of nia-
cin in persons with diabetes.
21-23
Most of the excess of serious musculoskeletal
adverse events with niacinlaropiprant was due
to myopathy. The absolute risk of myopathy in
the placebo group was much higher in China
than in Europe, and the relative risk with niacin
laropiprant versus placebo was 5.2 in China, as
compared with 1.5 in Europe. Consequently, as
reported previously,
11
the absolute excess of my-
opathy associated with adding niacinlaropiprant
to statin-based LDL cholesterollowering therapy
was more than 10 times as great among par-
ticipants in China as among those in Europe:
50 cases per 10,000 participants versus 3 cases
per 10,000 participants annually.
The observed excess of serious infections
an excess of 14 cases per 1000 participants as-
signed to receive niacinlaropiprant for 4 years
had not been expected. It is not possible to
determine the separate contributions of niacin
and laropiprant (with potential mechanistic expla-
nations for both drugs
24,25
) from the results of
HPS2-THRIVE alone. However, in the AIM-HIGH
trial, niacin alone was associated with a sig-
nificant increase in the risk of serious infection
(139 patients in the niacin group [8.1%] vs. 98
in the placebo group [5.8%], P = 0.008).
26
More-
over, the larger numbers of events of any sever-
ity attributed to infection in the AIM-HIGH
trial (674 participants in the niacin group [39.2%]
vs. 593 in the placebo group [35.0%]; P = 0.01)
provide an even more robust demonstration of
this hazard.
26
In HPS2-THRIVE, the unexpected excess of
serious bleeding events an excess of 7 cases
per 1000 participants treated with niacinlaro-
piprant for 4 years was distributed across
gastrointestinal, intracranial, and other sites.
Niacin is known to reduce platelet counts
27

(Table S4 in Supplementary Appendix 1) and
affect clotting,
28-31
but potential mechanisms
for bleeding have also been proposed for laro-
piprant.
32-34
In the smaller AIM-HIGH trial,
there were relatively few serious bleeding-related
events, and the excess with niacin alone was
not significant (59 participants in the niacin
group [3.4%] and 49 in the placebo group
[2.9%], P = 0.36).
26
However, the result is con-
sistent with the excess in HPS2-THRIVE, and
there was a significant excess in the larger
numbers of bleeding events of any severity re-
corded in the AIM-HIGH trial (174 participants
in the niacin group [10.1%] vs. 137 in the pla-
cebo group [8.1%], P = 0.04).
26
In conclusion, we evaluated the effects of
extended-release niacin combined with laropip-
rant, as compared with placebo, in 25,673 adults
with atherosclerotic vascular disease. Treat-
ment with extended-release niacinlaropiprant
did not significantly reduce the risk of major
vascular events but did significantly increase
the risk of serious adverse events. In light of
the consistency of the results with those from
previous trials of niacin alone, we believe that
the findings from HPS2-THRIVE are likely to
be generalizable to all high-dose niacin formu-
lations. Although niacin might still be relevant
for particular patient groups (e.g., patients at
high risk for vascular events who have high
levels of LDL cholesterol), any potential bene-
fits should be considered in the context of the
observed hazards.
Supported by grants from Merck, the U.K. Medical Research
Council, the British Heart Foundation, and Cancer Research
U.K. (to the University of Oxford), and by a grant from the Brit-
ish Heart Foundation Centre of Research Excellence, Oxford
(RE/08/004, to Dr. Hopewell).
Dr. Landray, Dr. Haynes, Dr. Hopewell, Dr. Parish, Dr. Aung,
Dr. Tomson, Mr. Wallendszus, Dr. Craig, Dr. Jiang, Dr. Collins,
and Dr. Armitage report receiving grant support from Merck. No
other potential conflict of interest relevant to this article was
reported.
Disclosure forms provided by the authors are available with
the full text of this article at NEJM.org.
We thank the participants in the study.
212
REFERENCES
1. Cholesterol Treatment Trialists (CTT)
Collaboration. Efficacy and safety of
more intensive lowering of LDL choles-
terol: a meta-analysis of data from
170,000 participants in 26 randomised
trials. Lancet 2010;376:1670-81.
2. Di Angelantonio E, Sarwar N, Perry P,
et al. Major lipids, apolipoproteins, and
risk of vascular disease. JAMA 2009;302:
1993-2000.
3. Carlson LA. Nicotinic acid: the broad-
spectrum lipid drug: a 50th anniversary
review. J Intern Med 2005;258:94-114.
4. Bays HE, Maccubbin D, Meehan AG,
Kuz netsova O, Mitchel YB, Paolini JF.
Blood pressure-lowering effects of extended-
release niacin alone and extended-release
niacin/laropiprant combination: a post hoc
analysis of a 24-week, placebo-controlled
trial in dyslipidemic patients. Clin Ther
2009;31:115-22.
5. Stone NJ, Robinson J, Lichtenstein
AH, et al. 2013 ACC/AHA guideline on
the treatment of blood cholesterol to re-
duce atherosclerotic cardiovascular risk
in adults: a report of the American Col-
lege of Cardiology/American Heart Asso-
ciation Task Force on Practice Guide-
lines. Circulation 2013 November 12 (Epub
ahead of print).
6. Catapano AL, Reiner Z, De Backer G,
et al. ESC/EAS guidelines for the manage-
ment of dyslipidaemias: the Task Force
for the management of dyslipidaemias of
the European Society of Cardiology (ESC)
and the European Atherosclerosis Society
(EAS). Atherosclerosis 2011;217:Suppl 1:S1-
S44. [Erratum, Atherosclerosis 2011;217:2.]
7. Jackevicius CA, Tu JV, Ko DT, de Leon
N, Krumholz HM. Use of niacin in the
United States and Canada. JAMA Intern
Med 2013;173:1379-81.
8. The AIM-HIGH Investigators. Niacin in
patients with low HDL cholesterol levels
receiving intensive statin therapy. N Engl
J Med 2011;365:2255-67. [Erratum, N Engl
J Med 2012;367:189.]
9. Lai E, De Lepeleire I, Crumley TM, et al.
Suppression of niacin-induced vasodilation
with an antagonist to prostaglandin D2
receptor subtype 1. Clin Pharmacol Ther
2007;81:849-57.
10. Maccubbin D, Bays HE, Olsson AG, et
al. Lipid-modifying efficacy and tolerability
of extended-release niacin/laropiprant in pa-
tients with primary hypercholesterolaemia
or mixed dyslipidaemia. Int J Clin Pract
2008;62:1959-70.
11. HPS2-THRIVE Collaborative Group.
HPS2-THRIVE randomized placebo-
controlled trial in 25 673 high-risk pa-
tients of ER niacin/laropiprant: trial de-
sign, pre-specified muscle and liver
outcomes, and reasons for stopping study
treatment. Eur Heart J 2013;34:1279-91.
12. Study of the Effectiveness of Addition-
al Reductions in Cholesterol and Homo-
cysteine (SEARCH) Collaborative Group.
Intensive lowering of LDL cholesterol with
80 mg versus 20 mg simvastatin daily in
12,064 survivors of myocardial infarction:
a double-blind randomised trial. Lancet
2010;376:1658-69. [Erratum, Lancet 2011;
377:126.]
13. LaRosa JC, Grundy SM, Waters DD, et
al. Intensive lipid lowering with atorva statin
in patients with stable coronary disease.
N Engl J Med 2005;352:1425-35.
14. Pedersen TR, Faergeman O, Kastelein
JJP, et al. High-dose atorvastatin vs usual-
dose simvastatin for secondary prevention
after myocardial infarction: the IDEAL
study: a randomized controlled trial. JAMA
2005;294:2437-45. [Erratum, JAMA 2005;
294:3092.]
15. Peto R, Pike MC, Armitage P, et al.
Design and analysis of randomized clini-
cal trials requiring prolonged observation
of each patient. II. Analysis and examples.
Br J Cancer 1977;35:1-39.
16. Yusuf S, Peto R, Lewis J, Collins R,
Sleight P. Beta blockade during and after
myocardial infarction: an overview of the
randomized trials. Prog Cardiovasc Dis
1985;27:335-71.
17. National Cholesterol Education Pro-
gram (NCEP) Expert Panel on Detection,
Evaluation, and Treatment of High Blood
Cholesterol in Adults (Adult Treatment
Panel III). Third Report of the National
Cholesterol Education Program (NCEP)
Expert Panel on Detection, Evaluation,
and Treatment of High Blood Cholesterol
in Adults (Adult Treatment Panel III) final
report. Circulation 2002;106:3143-421.
18. Voight BF, Peloso GM, Orho-Melander
M, et al. Plasma HDL cholesterol and risk
of myocardial infarction: a mendelian
randomisation study. Lancet 2012;380:
572-80. [Erratum, Lancet 2012;380:564.]
19. Clofibrate and niacin in coronary heart
disease. JAMA 1975;231:360-81.
20. Morgan JM, Capuzzi DM, Guyton JR.
A new extended-release niacin (Niaspan):
efficacy, tolerability, and safety in hyper-
cholesterolemic patients. Am J Cardiol
1998;82(12A):29U-34U, 39U-41U.
21. Goldberg RB, Jacobson TA. Effects of
niacin on glucose control in patients
with dyslipidemia. Mayo Clin Proc 2008;
83:470-8.
22. Shepherd JJ, Betteridge J, Van Gaal L,
European Consensus Panel. Nicotinic
acid in the management of dyslipidaemia
associated with diabetes and metabolic
syndrome: a position paper developed by
a European Consensus Panel. Curr Med
Res Opin 2005;21:665-82.
23. Fazio S, Guyton JR, Lin J, Tomassini
JE, Shah A, Tershakovec AM. Long-term ef-
ficacy and safety of ezetimibe/simvastatin
coadministered with extended-release ni-
acin in hyperlipidaemic patients with dia-
betes or metabolic syndrome. Diabetes
Obes Metab 2010;12:983-93.
24. Digby JE, Martinez F, Jefferson A, et
al. Anti-inflammatory effects of nicotinic
acid in human monocytes are mediated by
GPR109A dependent mechanisms. Arterio-
scler Thromb Vasc Biol 2012;32:669-76.
25. Ricciotti E, FitzGerald GA. Prosta-
glandins and inflammation. Arterioscler
Thromb Vasc Biol 2011;31:986-1000.
26. Anderson TJ, Boden WE, Desvigne-
Nickens P, et al. Safety profile of extended-
release niacin in the AIM-HIGH trial.
N Engl J Med 2014;371:288-90.
27. Guyton JR, Bays HE. Safety consider-
ations with niacin therapy. Am J Cardiol
2007;99(6A):22C-31C.
28. Nagakawa Y, Orimo H, Harasawa M.
The anti-platelet effect of niceritrol in pa-
tients with arteriosclerosis and the relation-
ship of the lipid-lowering effect to the anti-
platelet effect. Thromb Res 1985;40:543-53.
29. Serebruany V, Malinin A, Aradi D, Ku-
liczkowski W, Norgard NB, Boden WE.
The in vitro effects of niacin on platelet
biomarkers in human volunteers. Thromb
Haemost 2010;104:311-7.
30. Chesney CM, Elam MB, Herd JA, et al.
Effect of niacin, warfarin, and antioxidant
therapy on coagulation parameters in pa-
tients with peripheral arterial disease in
the Arterial Disease Multiple Intervention
Trial (ADMIT). Am Heart J 2000;140:631-6.
31. Philipose S, Konya V, Lazarevic M, et al.
Laropiprant attenuates EP3 and TP pros-
tanoid receptor-mediated thrombus for-
mation. PLoS One 2012;7(8):e40222.
32. Lai E, Schwartz JI, Dallob A, et al. Ef-
fects of extended release niacin/laropiprant,
laropiprant, extended release niacin and
placebo on platelet aggregation and bleed-
ing time in healthy subjects. Platelets
2010;21:191-8.
33. Lauring B, Dishy V, Luo WL, et al.
Laropiprant in combination with extended-
release niacin does not alter urine
11- dehydrothromboxane B2, a marker
of in vivo platelet function, in healthy,
hypercholesterolemic, and diabetic sub-
jects. J Clin Pharmacol 2009;49:1426-35.
34. Dallob A, Luo WL, Luk JM, et al. The
effects of laropiprant, a selective prosta-
glandin D
2
receptor 1 antagonist, on the
antiplatelet activity of clopidogrel or aspi-
rin. Platelets 2011;22:495-503.
Copyright 2014 Massachusetts Medical Society.

Nej Mo A 1300955

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Nej Mo A 1300955

Uploaded by

Copyright:

Available Formats

n engl j med 371;3 nejm.

org july 17, 2014

You might also like